Using CAPE to time market- timing?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
CULater
Posts: 1208
Joined: Sun Nov 13, 2016 10:59 am

Using CAPE to time market- timing?

Post by CULater » Sun Jan 28, 2018 1:24 pm

Lots of people are freaking out about the current high value of CAPE - 34.75 last I saw. We all know that you can't time the market using CAPE because CAPE doesn't predict short term market moves -- or does it? Here's what I'm wondering. Let's say that you are a disciple of market timing using moving averages. For example, a simple strategy is to be in the market when it is above the 10-month MA and go to cash when it is below the 10-month MA. One problem with MA strategies is that you can get whipsawed when the price repeatedly crosses above and below the MA.

But, how about this? Let's assume that the market is more likely enter a serious decline whenever CAPE is significantly elevated, as it appears to be now. If so, then going to cash when the market declines below it's MA might be a higher-probability strategy; i.e., more likely to work because the market regime is more favorable to market declines. We would have two indicators: high CAPE and a drop below the MA. This would allow me to stay in the market for now and ride it higher, if it continues to go up, and have a "rule" for going to cash if it declines.

According to this strategy, I would follow a MA "sell" rule only when CAPE is high. At other times, I would refrain from using MA timing. Is there any evidence that MA timing can be "timed" using CAPE? In other words, is the predictability of the MA contingent on the level of CAPE?
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

User avatar
Taylor Larimore
Advisory Board
Posts: 27198
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Re: Using CAPE to time market- timing?

Post by Taylor Larimore » Sun Jan 28, 2018 1:32 pm

CULater:

If there is a market-timing formula that worked it would soon become known and we would all be using it. Professor Burton Malkiel, in his classic "Random Walk Down Wall Street," put it this way:
Any regularity in the stock market that can be discovered and acted upon profitably is bound to destroy itself.
Listen to our mentor, Jack Bogle:
"After nearly 50 years in this business, I do not know of anybody who has done market timing successfully and consistently. I don't even know anybody who knows anybody who has done it successfully and consistently."
I once believed in market timing. I learned the hard way that it doesn't work. You can do better.

Best wishes.
Taylor
Last edited by Taylor Larimore on Sun Jan 28, 2018 1:45 pm, edited 2 times in total.
"Simplicity is the master key to financial success." -- Jack Bogle

CULater
Posts: 1208
Joined: Sun Nov 13, 2016 10:59 am

Re: Using CAPE to time market- timing?

Post by CULater » Sun Jan 28, 2018 1:39 pm

Taylor Larimore wrote:
Sun Jan 28, 2018 1:32 pm
CULater:

If there is a market-timing formula that worked it would soon become known and we would all be using it.

Listen to Mr. Bogle--not merchandiser's trying to sell you something:
"After nearly 50 years in this business, I do not know of anybody who has done market timing successfully and consistently. I don't even know anybody who knows anybody who has done it successfully and consistently."
Best wishes.
Taylor
I know. But we can dream can't we? I'll dream of a strategy that works because I'm lucky and I can call it wisdom... :greedy
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

lack_ey
Posts: 6583
Joined: Wed Nov 19, 2014 11:55 pm

Re: Using CAPE to time market- timing?

Post by lack_ey » Sun Jan 28, 2018 1:52 pm

Stock market history is not that long, especially for which we have okay data, and past patterns may not continue, especially regarding timing, especially now that transaction costs are lower.

What you're suggesting is more or less a valuation + trend following strategy, which is evaluated in this paper among others:
Market Timing: Sin a Little.

In the US valuation timing has been pretty miserable by most any measure, at least under common-sense rules. With the trend as your friend, maybe not as bad.

User avatar
nedsaid
Posts: 9888
Joined: Fri Nov 23, 2012 12:33 pm

Re: Using CAPE to time market- timing?

Post by nedsaid » Sun Jan 28, 2018 1:58 pm

I don't know, a 200 day moving CAPE average? I am being a bit facetious here. The best we can do is to try to lighten up on expensive asset classes and load up of cheaper asset classes. Problem is expensive can remain expensive and cheap can remain cheap for longer than what anyone dreamed possible. I admit to market timing in its mildest forms and I think many of us here have done so and not admitted to it. For the most part, I am a stay the course guy.

It would seem that tactical asset allocation based upon relative valuations would work. Problem is, fund companies have tried and failed with this approach. What looks good in theory doesn't seem to work out in real life.
A fool and his money are good for business.

magneto
Posts: 951
Joined: Sun Dec 19, 2010 10:57 am
Location: On Chesil Beach

Re: Using CAPE to time market- timing?

Post by magneto » Sun Jan 28, 2018 1:59 pm

CULater wrote:
Sun Jan 28, 2018 1:24 pm
But, how about this? Let's assume that the market is more likely enter a serious decline whenever CAPE is significantly elevated, as it appears to be now. If so, then going to cash when the market declines below it's MA might be a higher-probability strategy; i.e., more likely to work because the market regime is more favorable to market declines. We would have two indicators: high CAPE and a drop below the MA. This would allow me to stay in the market for now and ride it higher, if it continues to go up, and have a "rule" for going to cash if it declines.

According to this strategy, I would follow a MA "sell" rule only when CAPE is high. At other times, I would refrain from using MA timing. Is there any evidence that MA timing can be "timed" using CAPE? In other words, is the predictability of the MA contingent on the level of CAPE?
Favouring an Adaptive Value Investment approach with a momentum overlay, can well understand the line of thought.
Some comments :-

+ With a high CAPE, PE, low Dividend Yield, it is possible to see the market presently with limited upside potential and plentiful downside risk. Most investors can probably agree with that?
+ If the investor takes this view of the market, they might reduce the risk-side of AA proportionally, rather than seek to time.

+ 'Timing' in the sense of TIME employed in measurement, is relevant when looking at the Rate Of Change (ROC), first derivative (with respect to time), which is underlying logic hidden behind Moving Averages.
+ The 10 month crossover is popular, but crude, and may signal far too late, if the investor is hanging on to ride the wave to that last advantageous crest. Dow theory might be more useful, but whipsaws could proliferate.
+ The dedicated market-timer will want to look beyond ROC to the second derivative, then maybe drive themselves distracted with the third derivative, then the fourth and so on, ad infinitum !!!!

An Investment Plan might be a good idea?
Last edited by magneto on Sun Jan 28, 2018 2:03 pm, edited 1 time in total.
'There is a tide in the affairs of men ...', Brutus (Market Timer)

User avatar
Lauretta
Posts: 670
Joined: Wed Jul 05, 2017 6:27 am

Re: Using CAPE to time market- timing?

Post by Lauretta » Sun Jan 28, 2018 2:02 pm

I had looked at this question some time ago. Besides the paper referred to by lack_ey you might find also this of interest
https://alphaarchitect.com/2015/07/22/m ... -momentum/
I also corresponded with an academic who's regarded as an expert in the study of momentum and moving averages. He told me that he tested this strategy:
If CAPE < long-term average, do not use moving averages; otherwise, use the moving average timing strategy.
His conclusion was that this strategy performs worse than the strategy when moving average is used regardless of the value of CAPE.
When everyone is thinking the same, no one is thinking at all

User avatar
raven15
Posts: 327
Joined: Sun Nov 30, 2014 8:01 pm

Re: Using CAPE to time market- timing?

Post by raven15 » Sun Jan 28, 2018 2:03 pm

Don't to anything drastic. I thought that maybe you could set your bond percentage equal to CAPE during an annual or less often rebalance. So right now bonds would be at 35%.

That doesn't take into account bond yields though. Another might be to weight stocks and bonds according to epected returns: yield minus inflation for ten year treasury, 1/CAPE for stocks. .027-.02 + 1/35 = 3.5% .7%/3.5% = 20% bond allocation right now.

Or not.
It's Time. Adding Interest.

hilink73
Posts: 274
Joined: Tue Sep 20, 2016 3:29 pm

Re: Using CAPE to time market- timing?

Post by hilink73 » Sun Jan 28, 2018 2:06 pm

lack_ey wrote:
Sun Jan 28, 2018 1:52 pm
What you're suggesting is more or less a valuation + trend following strategy, which is evaluated in this paper among others:
Market Timing: Sin a Little.
I like the "disappointing" arrow on page 5. :D

JBTX
Posts: 3577
Joined: Wed Jul 26, 2017 12:46 pm

Re: Using CAPE to time market- timing?

Post by JBTX » Sun Jan 28, 2018 2:12 pm

Sometime back I downloaded the Shiller data in Excel. I was curious if there would be an optimal long term strategy using CAPE10.

While it is true that on average your returns will likely be lower when CAPE10 is high vs low, as a timing tool it isn't very good. For almost the entire period of mid 90's and beyond, it has been above average. So depending on your in and out strategy, you probably would have been out for most or all of the periods past 1991.

For pretty much all months past 1991 the CAPE was 19 or higher, until 2009. For 2009 it stayed in the 16-19 range for about a year, and rose back up above that ever since.

I tried coming up with optimal buy and sell strategy based on 100+ years, but you come up with ridiculous ones like sell at 32 and buy at 8. Of course if you use different time sets, you come up with dramatically different results.

If CAPE is very high, then I tend to adjust my AA down slightly. That is it.
Last edited by JBTX on Sun Jan 28, 2018 2:13 pm, edited 1 time in total.

ulrichw
Posts: 88
Joined: Sun Feb 03, 2013 12:17 am

Re: Using CAPE to time market- timing?

Post by ulrichw » Sun Jan 28, 2018 2:13 pm

CULater wrote:
Sun Jan 28, 2018 1:24 pm
[...] Let's assume that the market is more likely enter a serious decline whenever CAPE is significantly elevated, as it appears to be now. [...]
I'm glad you said "Let's assume", because this is an assumption that is not justified by the research behind CAPE (as I understand it, at least).

Shiller's work says there's a strong predictive value from CAPE for future returns over the next 10 or 20 years.

His work doesn't say anything about the next year's returns.

There are many ways that the market can deliver low returns over a decade-long timespan:
- There's your scenario: It could crash and then slowly recover
- It could rise sharply, and then crash, and repeat this cycle a few times
- It could stay flat
- It could rise sharply and consistently and end in a big crash

By measuring 10 or 20 year returns, all of these patterns would be treated the same, as long as the starting and end point are similar.

On your methodology: From brief indirect exposure to professional options trading during the years when I worked on financial trading systems (20 years ago), I'll tell you that the analytics these groups use are orders of magnitude more sophisticated than combining two moving averages (CAPE is effectively a moving average as it is essentially current price over the moving average of earnings).

These days there are groups no doubt applying machine learning algorithms over hundreds or thousands of parameters to try to predict the market.

It's very unlikely (in my opinion) that you'll have much success with something simplistic.

JBTX
Posts: 3577
Joined: Wed Jul 26, 2017 12:46 pm

Re: Using CAPE to time market- timing?

Post by JBTX » Sun Jan 28, 2018 2:18 pm

This is data I came up with back in 2010. I can't attest to its accuracy.

Hard to find a good trading strategy based upon this, if long term CAPE average is somewhere around 18 or 19 I think.

Notice post 2000 it never got below 21 even after the crash


1990.01 17.05
1990.02 16.51
1990.03 16.83
1990.04 16.81
1990.05 17.39
1990.06 17.82
1990.07 17.75
1990.08 16.17
1990.09 15.30
1990.1 14.82
1990.11 15.19
1990.12 15.85
1991.01 15.61
1991.02 17.35
1991.03 17.82
1991.04 18.16
1991.05 18.04
1991.06 18.02
1991.07 18.10
1991.08 18.51
1991.09 18.36
1991.1 18.35
1991.11 18.29
1991.12 18.44
1992.01 19.77
1992.02 19.58
1992.03 19.28
1992.04 19.30
1992.05 19.66
1992.06 19.32
1992.07 19.62
1992.08 19.72
1992.09 19.71
1992.1 19.37
1992.11 19.83
1992.12 20.45
1993.01 20.32
1993.02 20.55
1993.03 20.86
1993.04 20.46
1993.05 20.52
1993.06 20.61
1993.07 20.57
1993.08 20.81
1993.09 20.99
1993.1 21.11
1993.11 21.04
1993.12 21.17
1994.01 21.41
1994.02 21.26
1994.03 20.83
1994.04 20.06
1994.05 20.20
1994.06 20.29
1994.07 20.07
1994.08 20.54
1994.09 20.58
1994.1 20.40
1994.11 20.21
1994.12 19.91
1995.01 20.22
1995.02 20.80
1995.03 21.15
1995.04 21.64
1995.05 22.20
1995.06 22.72
1995.07 23.38
1995.08 23.28
1995.09 23.95
1995.1 23.93
1995.11 24.35
1995.12 25.03
1996.01 24.76
1996.02 25.98
1996.03 25.63
1996.04 25.43
1996.05 25.81
1996.06 25.97
1996.07 24.86
1996.08 25.41
1996.09 25.68
1996.1 26.48
1996.11 27.59
1996.12 27.72
1997.01 28.33
1997.02 29.27
1997.03 28.80
1997.04 27.59
1997.05 29.93
1997.06 31.26
1997.07 32.77
1997.08 32.59
1997.09 32.67
1997.1 32.90
1997.11 32.34
1997.12 33.03
1998.01 32.86
1998.02 34.71
1998.03 36.30
1998.04 37.28
1998.05 36.96
1998.06 36.80
1998.07 38.26
1998.08 35.42
1998.09 33.53
1998.1 33.77
1998.11 37.37
1998.12 38.82
1999.01 40.58
1999.02 40.40
1999.03 41.36
1999.04 42.71
1999.05 42.56
1999.06 42.18
1999.07 43.83
1999.08 41.93
1999.09 41.32
1999.1 40.55
1999.11 43.21
1999.12 44.20
2000.01 43.77
2000.02 42.19
2000.03 43.22
2000.04 43.53
2000.05 41.97
2000.06 42.78
2000.07 42.76
2000.08 42.87
2000.09 41.90
2000.1 39.37
2000.11 38.78
2000.12 37.28
2001.01 36.98
2001.02 35.84
2001.03 32.33
2001.04 32.17
2001.05 34.08
2001.06 33.07
2001.07 32.16
2001.08 31.40
2001.09 27.67
2001.1 28.58
2001.11 30.01
2001.12 30.50
2002.01 30.28
2002.02 29.09
2002.03 30.29
2002.04 29.01
2002.05 28.13
2002.06 26.39
2002.07 23.46
2002.08 23.59
2002.09 22.36
2002.1 21.95
2002.11 23.34
2002.12 23.10
2003.01 22.89
2003.02 21.21
2003.03 21.31
2003.04 22.42
2003.05 23.59
2003.06 24.83
2003.07 24.86
2003.08 24.64
2003.09 25.24
2003.1 25.68
2003.11 25.94
2003.12 26.63
2004.01 27.65
2004.02 27.65
2004.03 26.88
2004.04 26.90
2004.05 25.90
2004.06 26.40
2004.07 25.69
2004.08 25.17
2004.09 25.66
2004.1 25.41
2004.11 26.46
2004.12 27.14
2005.01 26.58
2005.02 26.74
2005.03 26.33
2005.04 25.40
2005.05 25.64
2005.06 26.06
2005.07 26.28
2005.08 26.10
2005.09 25.72
2005.1 24.87
2005.11 25.92
2005.12 26.43
2006.01 26.46
2006.02 26.24
2006.03 26.32
2006.04 26.14
2006.05 25.64
2006.06 24.74
2006.07 24.69
2006.08 25.04
2006.09 25.63
2006.1 26.53
2006.11 26.92
2006.12 27.27
2007.01 27.20
2007.02 27.31
2007.03 26.22
2007.04 26.97
2007.05 27.54
2007.06 27.41
2007.07 27.40
2007.08 26.14
2007.09 26.72
2007.1 27.31
2007.11 25.72
2007.12 25.95
2008.01 24.01
2008.02 23.49
2008.03 22.60
2008.04 23.35
2008.05 23.69
2008.06 22.41
2008.07 20.90
2008.08 21.39
2008.09 20.36
2008.1 16.38
2008.11 15.25
2008.12 15.37
2009.01 15.17
2009.02 14.12
2009.03 13.32
2009.04 14.98
2009.05 15.99
2009.06 16.38
2009.07 16.69
2009.08 18.09
2009.09 18.83
2009.1 19.35
2009.11 19.81
2009.12 20.32
2010.01 20.52
2010.02 19.91
2010.03 21.00
2010.04 21.80
2010.05 20.47
2010.06 19.73
2010.07 19.66
2010.08 19.77
2010.09 20.37
2010.1 21.03

CULater
Posts: 1208
Joined: Sun Nov 13, 2016 10:59 am

Re: Using CAPE to time market- timing?

Post by CULater » Sun Jan 28, 2018 2:21 pm

Lauretta wrote:
Sun Jan 28, 2018 2:02 pm
I had looked at this question some time ago. Besides the paper referred to by lack_ey you might find also this of interest
https://alphaarchitect.com/2015/07/22/m ... -momentum/
I also corresponded with an academic who's regarded as an expert in the study of momentum and moving averages. He told me that he tested this strategy:
If CAPE < long-term average, do not use moving averages; otherwise, use the moving average timing strategy.
His conclusion was that this strategy performs worse than the strategy when moving average is used regardless of the value of CAPE.
Can you explain the underlined statement you made? Not sure I am following.
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

CULater
Posts: 1208
Joined: Sun Nov 13, 2016 10:59 am

Re: Using CAPE to time market- timing?

Post by CULater » Sun Jan 28, 2018 2:30 pm

lack_ey wrote:
Sun Jan 28, 2018 1:52 pm
Stock market history is not that long, especially for which we have okay data, and past patterns may not continue, especially regarding timing, especially now that transaction costs are lower.

What you're suggesting is more or less a valuation + trend following strategy, which is evaluated in this paper among others:
Market Timing: Sin a Little.

In the US valuation timing has been pretty miserable by most any measure, at least under common-sense rules. With the trend as your friend, maybe not as bad.
Yes. Exactly what I was thinking: combining a valuation strategy with a momentum (trend-following). I see that the paper identifies the problem of drifting valuation measures, such as CAPE. It also suggest that this strategy has marginal utility. However, I was interested in the comment that adding a momentum component might have a psychological benefit of helping investors become more patient long-term value investors. What do you think of that notion?
May you have the hindsight to know where you've been, The foresight to know where you're going, And the insight to know when you've gone too far. ~ Irish Blessing

User avatar
Lauretta
Posts: 670
Joined: Wed Jul 05, 2017 6:27 am

Re: Using CAPE to time market- timing?

Post by Lauretta » Sun Jan 28, 2018 2:33 pm

CULater wrote:
Sun Jan 28, 2018 2:21 pm
Lauretta wrote:
Sun Jan 28, 2018 2:02 pm
I had looked at this question some time ago. Besides the paper referred to by lack_ey you might find also this of interest
https://alphaarchitect.com/2015/07/22/m ... -momentum/
I also corresponded with an academic who's regarded as an expert in the study of momentum and moving averages. He told me that he tested this strategy:
If CAPE < long-term average, do not use moving averages; otherwise, use the moving average timing strategy.
His conclusion was that this strategy performs worse than the strategy when moving average is used regardless of the value of CAPE.
Can you explain the underlined statement you made? Not sure I am following.
Yes I didn't express this correctly; I mean that using MA only as a timing indicator was better than adding a criterion based on CAPE to decide whether to be in or out of stocks.
So his tests showed that using trend following only was better than combining it with CAPE.
This is because the stock market has secular trends. So the CAPE reaches its top, then it
decreases to the long-term average, and then continues to decrease to a bottom point. If one invests right after the CAPE crosses its long-term average, during some period (that can be rather long) they lose money on stocks. The moving average strategy, on the other hand, protects to some degree from losses during a long secular bear market as you reinvest only when the trend turns positive.
When everyone is thinking the same, no one is thinking at all

JBTX
Posts: 3577
Joined: Wed Jul 26, 2017 12:46 pm

Re: Using CAPE to time market- timing?

Post by JBTX » Sun Jan 28, 2018 2:55 pm

FWIW this guys reasearch says shorter term like 200 day MA don’t work.

https://adamhgrimes.com/200-day-moving-average-work/

I suspect these like other things may work, until they don’t.

One issue with moving averages is whipsaw. If you get in a period where the average moves around the MA line above and below, and you end up buying high and selling low over and over. The effectiveness of a moving average assumes long term trend moves, which isn't necessarily always the case.

A significant driver of high Cape10 is low interest rates. Looking at cape10 or moving averages isn’t going to tell you what may happen structurally with interest rates.

User avatar
willthrill81
Posts: 4982
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Using CAPE to time market- timing?

Post by willthrill81 » Sun Jan 28, 2018 5:33 pm

Taylor Larimore wrote:
Sun Jan 28, 2018 1:32 pm
If there is a market-timing formula that worked it would soon become known and we would all be using it.
I've heard this often and considered it. I do not believe it to be true.

There are several asset classes that have, historically, outperformed the total stock market (e.g. small cap value). This is very well known, yet the overwhelming majority of dollars going into stocks are not going into SCV. Why? Because (1) participants don't believe that it will continue to outperform, (2) they believe that SCV's outperformance is limited to a handful of years, (3) they believe that their risk-adjusted returns will be higher with other asset classes, or (4) they do not have easy access to invest in SCV (e.g. in their 401k). Other reasons likely exist.

The same could be said of market-timing. Even though it has been shown by academics that many trend following strategies that were devised nearly a century ago have indeed outperformed the market, they have also undergone significant periods of time in which they lagged the market's returns. Investors, as a group, do not tend to be very patient and have a strong urge to chase returns.

So no, even if a particular strategy did work in the past, everybody would not be using it going forward.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

JBTX
Posts: 3577
Joined: Wed Jul 26, 2017 12:46 pm

Re: Using CAPE to time market- timing?

Post by JBTX » Sun Jan 28, 2018 6:01 pm

willthrill81 wrote:
Sun Jan 28, 2018 5:33 pm
Taylor Larimore wrote:
Sun Jan 28, 2018 1:32 pm
If there is a market-timing formula that worked it would soon become known and we would all be using it.
I've heard this often and considered it. I do not believe it to be true.

There are several asset classes that have, historically, outperformed the total stock market (e.g. small cap value). This is very well known, yet the overwhelming majority of dollars going into stocks are not going into SCV. Why? Because (1) participants don't believe that it will continue to outperform, (2) they believe that SCV's outperformance is limited to a handful of years, (3) they believe that their risk-adjusted returns will be higher with other asset classes, or (4) they do not have easy access to invest in SCV (e.g. in their 401k). Other reasons likely exist.

The same could be said of market-timing. Even though it has been shown by academics that many trend following strategies that were devised nearly a century ago have indeed outperformed the market, they have also undergone significant periods of time in which they lagged the market's returns. Investors, as a group, do not tend to be very patient and have a strong urge to chase returns.

So no, even if a particular strategy did work in the past, everybody would not be using it going forward.
Since 1998, which was beginning of time for chart, vanguard small cap value basically is same as small cap growth. Yes, they beat the S&P, but that can probably be explained by risk reward.

It could very well be that the advantage that once existed no longer does.

http://quotes.morningstar.com/chart/fun ... ture=en-US

staythecourse
Posts: 5811
Joined: Mon Jan 03, 2011 9:40 am

Re: Using CAPE to time market- timing?

Post by staythecourse » Sun Jan 28, 2018 6:14 pm

Question back to the OP...

What makes you think this hasn't already been thought of by somebody else? If all the data is available to make the calculations for each metric then what do you think the high end computers at the hedge funds are doing 24/7?

Either you think they have not thought about it yet? Unlikely to zero chance. Or they don't know how to implement it correct and somehow you and only you know? Or they do know and you have stumbled onto their magic black box?

All the above is possible (not likely though), but one thing is certain this idea has already been thought of and my guess would be about 10-20 years ago.

Good luck.

p.s. More and more I hear of valuation metrics the more I think about the story of the finance professor and the student and the $20 bill lying on the street.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

stlutz
Posts: 4606
Joined: Fri Jan 02, 2009 1:08 am

Re: Using CAPE to time market- timing?

Post by stlutz » Sun Jan 28, 2018 6:40 pm

I think there are two approaches to looking at market valuation levels that get mixed up in these threads and result in thinking that is problematic:

a) OMG! The price-X ratio is about where it was at some other key market peaks. The market is about to crash!

b) When valuation ratios are higher, you get less returns from the earnings yield. Absent any change in valuations going forward, one should expect lower future returns because of this and because the historical returns included an increase in valuation ratios. Note that this group isn't predicting a crash, they are just looking at mathematically likely calculations.

The former is used as a short-to-intermediate term market timing signal; the later factors into one's evaluation of their need and willingness to take risk--e.g. if the likely rewards from taking risk won't be as great over the long haul, maybe it's not worth it to take as much? Maybe or maybe not--depends on your personal situation.

Moving average strategies don't consistently provide superior absolute returns. They to do tend to provide superior risk-adjusted returns provided one is doing it in a tax-sheltered account. They particularly help deal with tail risk, although they can create tail risk of their own (think of a choppy market where the buy-and-hold investor breaks even but the MA market timer loses 20% because of whipsaws).

Other ways that can be used to improve one's risk adjusted returns are to a) pick the best stocks (good luck with that) or b) hold a diversified portfolio of stocks and bonds. Anecdotally, it seems that this group actually performs the best and actually sticks with their strategy (by comparing the number of pretty darn rich people on this board versus places where market timing is discussed).

The problem with market timers in general is that few of them actually stick with their approach for decades. Instead, they see a case where it recently would have worked very well (e.g. selling stocks early in 2008), they adopt the strategy, execute a number of trades that cause them to perform worse than B&H, and then they drop the strategy

User avatar
willthrill81
Posts: 4982
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: Using CAPE to time market- timing?

Post by willthrill81 » Sun Jan 28, 2018 8:33 pm

JBTX wrote:
Sun Jan 28, 2018 6:01 pm
willthrill81 wrote:
Sun Jan 28, 2018 5:33 pm
Taylor Larimore wrote:
Sun Jan 28, 2018 1:32 pm
If there is a market-timing formula that worked it would soon become known and we would all be using it.
I've heard this often and considered it. I do not believe it to be true.

There are several asset classes that have, historically, outperformed the total stock market (e.g. small cap value). This is very well known, yet the overwhelming majority of dollars going into stocks are not going into SCV. Why? Because (1) participants don't believe that it will continue to outperform, (2) they believe that SCV's outperformance is limited to a handful of years, (3) they believe that their risk-adjusted returns will be higher with other asset classes, or (4) they do not have easy access to invest in SCV (e.g. in their 401k). Other reasons likely exist.

The same could be said of market-timing. Even though it has been shown by academics that many trend following strategies that were devised nearly a century ago have indeed outperformed the market, they have also undergone significant periods of time in which they lagged the market's returns. Investors, as a group, do not tend to be very patient and have a strong urge to chase returns.

So no, even if a particular strategy did work in the past, everybody would not be using it going forward.
Since 1998, which was beginning of time for chart, vanguard small cap value basically is same as small cap growth. Yes, they beat the S&P, but that can probably be explained by risk reward.

It could very well be that the advantage that once existed no longer does.

http://quotes.morningstar.com/chart/fun ... ture=en-US
From 1999-2017, the Sharpe ratio for VISVX was .52, compared to .39 for VTSMX. So it outperformed in both absolute and risk-adjusted returns.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
grayfox
Posts: 4927
Joined: Sat Sep 15, 2007 4:30 am

Re: Using CAPE to time market- timing?

Post by grayfox » Sun Jan 28, 2018 9:35 pm

CULater wrote:
Sun Jan 28, 2018 1:24 pm
Lots of people are freaking out about the current high value of CAPE - 34.75 last I saw. We all know that you can't time the market using CAPE because CAPE doesn't predict short term market moves -- or does it? Here's what I'm wondering. Let's say that you are a disciple of market timing using moving averages. For example, a simple strategy is to be in the market when it is above the 10-month MA and go to cash when it is below the 10-month MA. One problem with MA strategies is that you can get whipsawed when the price repeatedly crosses above and below the MA.

But, how about this? Let's assume that the market is more likely enter a serious decline whenever CAPE is significantly elevated, as it appears to be now. If so, then going to cash when the market declines below it's MA might be a higher-probability strategy; i.e., more likely to work because the market regime is more favorable to market declines. We would have two indicators: high CAPE and a drop below the MA. This would allow me to stay in the market for now and ride it higher, if it continues to go up, and have a "rule" for going to cash if it declines.

According to this strategy, I would follow a MA "sell" rule only when CAPE is high. At other times, I would refrain from using MA timing. Is there any evidence that MA timing can be "timed" using CAPE? In other words, is the predictability of the MA contingent on the level of CAPE?
This idea was discussed years ago on this forum. c. 2008-2010.

magneto
Posts: 951
Joined: Sun Dec 19, 2010 10:57 am
Location: On Chesil Beach

Re: Using CAPE to time market- timing?

Post by magneto » Mon Jan 29, 2018 4:36 am

Lauretta wrote:
Sun Jan 28, 2018 2:33 pm
Yes I didn't express this correctly; I mean that using MA only as a timing indicator was better than adding a criterion based on CAPE to decide whether to be in or out of stocks.
So his tests showed that using trend following only was better than combining it with CAPE.
This is because the stock market has secular trends. So the CAPE reaches its top, then it
decreases to the long-term average, and then continues to decrease to a bottom point. If one invests right after the CAPE crosses its long-term average, during some period (that can be rather long) they lose money on stocks. The moving average strategy, on the other hand, protects to some degree from losses during a long secular bear market as you reinvest only when the trend turns positive.
The secular trends bit is a key point.
Those expecting CAPE or the other market measures to cross the long-term average anytime soon, may well be disappointed.
Markets have tended to swing with the Juglar (intermediate) wave (about 9 years), superimposed on the Krondatiev (long) wave (about 50 years).
CAPE can be considered much like a ship swinging at anchor; but every so often that anchor drags.
Medium-term we may be more interested in that intermediate wave and the crossing of the medium-term average.
There is no need to believe in predictive powers for cycles, or believe in cycles per se, merely to note a curious and irrritating tendency for them to exist; rather than for things to stay steady and constant, as in a perfect world.
Last edited by magneto on Mon Jan 29, 2018 1:46 pm, edited 1 time in total.
'There is a tide in the affairs of men ...', Brutus (Market Timer)

User avatar
just frank
Posts: 1455
Joined: Sun Nov 02, 2014 4:13 pm
Location: Philly Metro

Re: Using CAPE to time market- timing?

Post by just frank » Mon Jan 29, 2018 5:13 am

Schiller has been clear that CAPE is not useful for timing. Why wouldn't you believe him?

A lot of traders use estimated forward PE, not CAPE. The market jumped because forward earnings estimates improved, to keep the same forward PE. Currently more like 18, less scary than 34.

This:
http://fundamentalis.com/?p=7531

Post Reply